中文摘要：

属性约简是数据挖掘之中最核心的问题，是任何一个部门决策知识获取的关键技术。基于深入研究模糊粗糙理论、直觉模糊粗糙集理论在属性约简知识方面的研究成果，通过定义区间模糊粗糙集的正域、依赖度与非依赖度等相关概念，提出一种启发式区间直觉模糊粗糙集属性约简方法。结果表明：该方法在知识约简中是可行的，并且相比差别矩阵方法，能有效降低空间和时间复杂度。

英文摘要：

Attribute reduction is the most crucial problem among data mining ,which is the key technology of decision‐making knowledge acquisition .This paper studies attribute reduction theory of fuzzy rough sets theory and intuitionistic fuzzy rough sets theory .By defining the interval fuzzy rough sets , their positive fields , dependency and non‐dependency and other related concepts ,it propose a heuristic interval intuitionistic fuzzy rough set attribute reduction method .The results show that the method in knowledge reduction is feasible , which can effectively reduce space and time complexity , compared with the method of discernibility matrix .